The choice of a cutting score for criterion-related tests influences decisions related to classifying people into dichotomous categories. This paper proposes an empirical methodology for determining the best cutting score when there is information about the test score frequency distribution of test-takers defined as actually successful and actually unsuccessful on some criterion. The method is based on two statistics calculated for each possible cutting score. The first is a pure hit rate, representing the proportion of correct classifications above those expected by chance. Second is a chi-square statistic for testing the significance of the difference between the population frequencies of the two types of misclassifications errors. A cutting score summary table is developed based on the information about the test score frequency distributions of two validation samples based on actually successful and actually unsuccessful samples. Cutting scores are divided into those that yield equal frequencies of the two types of misclassification errors and those in which the frequency of one type of error is higher than that of the other. The cutting score summary table facilitates the determination of the best cutting score in each category. (Contains 2 tables and 19 references.) (SLD)